Because plasma lipoprotein(a) (Lp(a)) is atherogenic and apolipoprotein(a) (apo(a)) isoform size may affect plasma Lp(a) concentrations, we evaluated a new, lectin affinity-based method for serum Lp(a)-cholesterol (Lp(a)-C), then studied the relationship between apo(a) isoforms, total Lp(a) mass, Lp(a)-C, and other serum lipid parameters. For apo(a) phenotyping, the isoforms were classified according to low (less than or equal to 20), medium (21-29), and high (greater than or equal to 30) number of kringle 4 repeats. While no detectable loss of Lp(a) occurred in the initial filtrate and wash buffer of the LP(a)-C method, an Lp(a) concentration-dependent fraction (less than or equal to 10 percent of total Lp(a) mass, less than or equal to 25 percent of total Lp(a)-C] was often obtained by repeating the elution step, requiring addition to the Lp(a)-C results. Serum Lp(a) only moderately correlated with Lp(a)-C (r-squared=0.73). The regression equation (Lp(a)-C = 0.14 * Lp(a) + 24.2) indicated that Lp(a)-C as a fraction of total Lp(a) mass decreases with increasing Lp(a) concentrations. Further, decreasing apo(a) isoform sizes tended to increase the Lp(a)-C and, to a different degree, the total Lp(a) mass concentrations, indicating that Lp(a)-C is, at least partially, an Lp(a) mass-independent measure of Lp(a) particles. No other lipid parameters (glycerol-blanked triglycerides, total cholesterol, high-density lipoprotein-cholesterol, apolipoproteins A-I and B, low-density lipoprotein-cholesterol) showed a similar segregation with apo(a) phenotypes.A new, quantitative approach based on image analysis of restriction endonuclease digested polymerase chain reaction'amplified genomic DNA products was developed for apolipoprotein E genotyping. Quantitation of band intensity considerably improved the diagnostic reliability of apo E genotyping for both lipid disorders and Alzheimer's disease.In collaborative clinical studies, we continued to study the association between apo(a) isoforms and premature thromboembolic events in patients with systemic lupus erythematosus and confirmed that knowledge of apo(a) phenotypes improves prediction.